[Stoves] Tar characterization

[Bond, Tami C]

Hi Paul, and all,

Sorry that I have not been keeping up with Stoves discussions….

We have fussed a bit with trying to characterize the organic matter emitted from wood combustion.

What kind of characterization do you want? Full speciation is well-nigh impossible. But one needs to start with the question of what you want to know and why?
The problem is that this stuff is complex chemically and doesn’t take well to matching with standards.

Brief summary of what I know:
- GC/MS is really hard for characterization because it is poor at identifying compounds that are large (molecular weight) and polar. Wood organics are both. You’ll get the light organic material, but that’s a very small fraction of the emissions. You can get into derivatization of the emissions before GC/MS and that may get you a little more. See all Jamie Schauer’s, Wolfgang Rogge’s work. Most of it ends up being classified as “UCM” (unresolved complex mixture).
- FTIR— have tried a little. Although it nominally gives you functional groups, and occasionally you can see differences between one kind of wood or burning condition and another, it’s not very quantitative. Plus there are a lot of overlaps between the interesting functional groups, so you can’t distinguish some of them.
- 1NMR— kind of like FTIR— you can sort of see functional groups and that’s about it.
- UV-Vis— used for humic matter (and some “tar” isn’t far off that)— Again not quantitative, rather just gives a sense of how strong an absorber it is. You won’t see any structure as you do in UV-Vis of individual compounds. There are so many compounds that the entire spectrum smooths out.
    —> From our UV-Vis work we found that most of the stuff dissolves in methanol, not hexane not water. (Chen & Bond, Atmos Chem Physics 2010, open access)
- 13CNMR— which can get at the carbon skeleton rather than functional groups… tough. You need HUGE sample sizes and they don’t stay suspended because anything concentrated enough to measure wants to precipitate. We ended up using DCM, I think— I’d have to check. And then, the result was not all that interesting: we again couldn’t distinguish the unsaturated bonds, e.g. aromatic vs conjugated but linear bonds.
- LC/MS— expensive and full of artifacts— I’ve never gotten into it. Too expensive and haven’t seen the benefit, although I’d be interested if someone tried.

The question, again, is what you want to know and why. If it’s some kind of physical property of the emissions, you’d be better off measuring that.

Tami



On Sep 2, 2016, at 12:22 PM, Paul Medwell <paul.medwell at adelaide.edu.au<mailto:paul.medwell at adelaide.edu.au>> wrote:

Dear Philip,

Thanks for another great suggestion.

This is also something that is planned. As part of an unrelated project we have been looking at this: we are still in the process of finding a solvent that has a collection yield, that doesn't interfere with (or gives an interference profile that we can correct for) and that OH&S will let us have. Fortunately, for the cookstove work there are fewer constraints than the other project (which is limited to sampling extremely low flowrates). We'll find a solution and share it.

Cheers,
Paul


-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.bioenergylists.org/pipermail/stoves_lists.bioenergylists.org/attachments/20161112/4e4384d5/attachment.html>